Experimental site: The experiment was conducted at the field laboratory of the Department of Genetics and Plant Breeding, Bangladesh Agricultural University (BAU), Mymensingh under irrigated condition. The experiment was carried out in the land which was medium high belonging to the old Brahmaputra floodplain, Agro-Ecological Zone (AEZ-9) having non calcareous dark gray soil (UNDP and FAO, 1998). The soil was sandy loam with pH value of 6.5. The climate of the experimental location during the wheat crop season is characterized by a relatively low rainfall and temperature from December to March.
Plant materials The materials of this study were consisted of 20 genotypes of spring wheat. They were collected from Wheat germplasm collection in the Department of Genetics and Plant Breeding, BAU, Mymensingh.
Cultivation procedure The experimental plot was thoroughly prepared by ploughing with power tiller followed by harrowing and laddering. After laddering the weeds and the stubbles of previous crops were removed from the land. Recommended doses of urea, tripple super phosphate, muriate of potash, gypsum and cowdung were applied at 200, 150, 80, 50 and 10000 kg ha-1, respectively. Half the amounts of urea, full amount of triple superphosphate, muriate of potash, gypsum and cowdung were applied to the soil at the time of final land preparation. The rest of the urea was top dressed in two equal splits one at tillering and other at booting stages. The experiment was set up in a randomized complete block design with three replications. The plot size was 2 m - 2.5 m. The distance between two plots was 50 cm and the distance between two blocks was 100 cm. The spacing between row to row and plant to plant in the same row were 20 cm and 5 cm, respectively. The seeds of 20 wheat genotypes were sown in the field following line sowing. Intercultural practices were done uniformly for all genotypes. Thinning was done 25 days after sowing. Weeding was done when become necessary. The crop was irrigated two times one at the crown root initiation stage and the other at the peak tillering stage. Ten plants from each plot were randomly selected to collect the data and they were harvested by uprooting. Border plants were discarded to avoid border effects. Data were collected from 10 randomly selected plants for measuring the growth, yield and yield contributing characters of wheat genotypes.
Statistical Analysis: All the collected data were statistically analyzed. Mean values for each character were used for statistical analysis. Analysis of variance was done for all the characters under study using the mean values (Singh and Chaudhury, 1985). The total variability was partitioned into treatments (genotypes), blocks (replications) and error components. When the F-ratio was significant, the differences between mean values were tested by Duncan's Multiple Range Test (DMRT) Steel and Torrie (1960). Genotypic and phenotypic variances, coefficients of variation, heritability, and genetic advance were estimated according to the formula given by Johnson et al, (1955). Moreover, genetic advance in percent of mean was calculated by the following formula proposed by Comstock and Robinson (1952), genotypic and phenotypic covariances were calculated using the formula suggested by Singh and Chaudhary (1985), genotypic and phenotypic correlation coefficients were estimated using the following formula suggested by Johnson et al. (1955) and Singh and Chaudhury (1985).
The components of correlation coefficients of different yield contributing characters with yield were partitioned into components of direct and indirect effects by path coefficient analysis. Path coefficient analysis was done according to the procedure stated by Singh and Chaudhury (1985) and Dabholkar (1992) which was originally suggested by Wright (1921). In the present study, grain yield plant-1 was considered as the resultant character (effect) and the seven yield contributing characters were considered as the causal factors. After calculating the direct and indirect effects of the characters, residual effect (R) was calculated using the following formula suggested by Singh and Choudhury (1985). Selection indices were constructed using the methods developed by Smith (1936) based on the discriminant function as proposed by Fisher (1936). The expected genetic advance (GA) through selection was calculated by formula.